Wide-Angle Camera Calibration for Accurate Grid Pixel Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing storage systems face challenges in accurately determining the location of load handling devices operating remotely within a grid framework structure, particularly when communication is lost or devices become unresponsive, leading to potential collisions and misalignment.
Innovation Solution
A method and system for calibrating wide-angle or ultra wide-angle cameras above the grid framework, using a neural network to process images and map distorted pixels to correct grid coordinates, updating parameters to enhance accuracy, and storing these for precise image-to-grid mapping.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If wide-angle or ultra wide-angle cameras are used to monitor the grid framework structure, then the field of view and coverage area are improved, but image distortion increases making accurate pixel-to-grid mapping difficult
Solution Approach 1:
The system changes the parameters of the camera mapping by using multiple parameters (focal length, tilt angle, rotation angle, height) to characterize the camera's position and orientation. By adjusting and optimizing these parameters, the system achieves accurate pixel-to-grid mapping despite the wide-angle distortion, resolving the contradiction between wide field of view and mapping precision.
2Ease of operation
If manual calibration methods are used to determine camera parameters, then the process is simple to understand, but the accuracy and reliability of parameter determination is insufficient
Solution Approach 1:
The system replaces manual mechanical calibration methods with an automated computer vision approach. A neural network automatically processes images to detect grid lines and calculate camera parameters, substituting the manual mechanical adjustment process with an automated digital system that achieves both high accuracy and reliability while maintaining operational simplicity.
3Ease of manufacture
If traditional calibration approaches are used, then the system is easier to implement, but the reliability of load handling device location determination deteriorates when communication is lost
Solution Approach 1:
The system implements a feedback mechanism where the calibrated camera continuously captures images of the grid framework, and the neural network processes these images to determine the positions of load handling devices. This visual feedback loop provides redundant location information that maintains reliability even when communication with the load handling devices is lost, as the system can independently track device positions through image analysis.
4Measurement precision
If multiple parameters are used for camera mapping, then the mapping accuracy is improved, but the computational complexity and calibration time increase
Solution Approach 1:
The system performs preliminary calibration to determine the camera parameters (focal length, tilt angle, rotation angle, height) before actual operation. By pre-calculating and storing these parameters, the system avoids complex real-time computations during operation, thus achieving high mapping accuracy without excessive computational complexity during normal use.
Data Source
AI summary
A method and system for calibrating a wide-angle or ultra wide-angle camera disposed above a grid of a storage system. An image of a grid section and initial values of a plurality of parameters corresponding to the camera are obtained. The image is processed, using a neural network trained to detect/predict sets of parallel tracks in images of grid sections captured by wide-angle or ultra wide-angle cameras, to generate a model of the sets of parallel tracks as captured in the image. Selected pixels in the model are mapped to corresponding points on the grid using a mapping based on the plurality of parameters, with the initial values used as inputs to the mapping. An error function is determined based on a discrepancy between mapped grid coordinates of the points and known grid coordinates. The initial values of the parameters are updated based on the error function and the updated values are stored for mapping pixels in images of the grid section captured by the camera to corresponding points on the grid of the storage system.